一种基于ICA的同态盲反卷积算法
|
摘要
盲反卷积是图像处理、语音信号处理、通信、系统辨识和声学等许多研究和应用的基本问题,具有重要的理论与应用价值。根据无损检测中盲反卷积问题的特点,提出了一种新的基于ICA的同态盲反卷积算法。该算法首先将检测信号变换到复倒谱域,将卷积混合模型变为线性混合模型,即ICA问题;然后通过ICA将系统冲击响应和输入信号分离;最后,根据分离的复倒谱信号,重构其时域信号。论文提出的盲反卷积算法具有运算量小,计算速度快,分离精度高等特点,且不受信道是否为最小相位信道的影响。计算机模拟和实验数据都证明了算法的有效性。
Blind deconvolution(BD) is a fundamental problem in image processing,speech signal processing,communi-cation,system identification and so on,which has attracted extensive research interests in the past decade.In a traditional homomorphic deconvolution system,linear filtering is often used.In this paper,the author presents a novel blind deconvolution algorithm,in which independent component analysis(ICA) instead of linear filtering is applied to separate the complex cepstrum signals.The algorithm provided is fast and capable of separating the signals with high precision.Simulation and experimental results demonstrate the method for blind deconvolution is effective.
Blind deconvolution(BD) is a fundamental problem in image processing,speech signal processing,communi-cation,system identification and so on,which has attracted extensive research interests in the past decade.In a traditional homomorphic deconvolution system,linear filtering is often used.In this paper,the author presents a novel blind deconvolution algorithm,in which independent component analysis(ICA) instead of linear filtering is applied to separate the complex cepstrum signals.The algorithm provided is fast and capable of separating the signals with high precision.Simulation and experimental results demonstrate the method for blind deconvolution is effective.
引文
1.刘喜武,刘洪.地震盲反褶积综述[J].地球物理学进展,2003;18(2):203～209
2.Subram aniam S,Petropulu A P,W endt C.Cepstrum-based deconvolu-tion for speech dereverberation[J].IEEE Transactions on Speech and Audio Processing,1996;4(5):392～396
3.Kasprzak W.Blind source deconvolution by hom om orphic filtering in Fourier space[C].In:Proceedings of IAPR ICCVG2002,Zakopane,Poland,2002-09:431～436
4.Lupini A,Solazzi M,Piazza F et al.Frequency domain m ultichannel blind deconvolution using adaptive spline functions[C].In:Proceedings of IEEEINNS-ENNS International Joint Conference on Neural Net-works IJCNN2001,W ashington,USA,2001-07;1:464～468
5.Torkkola K.Blind separation for audio signals-Are we there yet?[C].In:Proceedings of1st International W orkshop on Independent Com-ponent Analysis and Blind Signal Separation,Aussois,France,1999:239～244
6.Sabala I,Cichoki A,Am ari S.Relationships between instantaneous blind source separation and m ultichannel blind deconvolution[C].In:Proceedings of Intl.Joint Conf.on Neural Networks(IJCNN'98),An-chorage,AK,1998-05;1:39～44
7.Yang H H.On-line blind equalization via on-line blind separation[J].Signal Processing,1998;68(3):271～281
8.Zhang Y,Kassam S A.Blind Separation and equalization using frac-tional sam pling of digital com m unication signals[J].Signal Processing,2001;81:2591～2608
9.Enescu M,Zhang Y L,Kassam S A et al.Recursive estim ator for blind M IM O equalization via BSS and fractional sam pling[C].In:Proceedings of Third IEEE Signal Processing W orkshop on Signal Processing Ad-vances in W ireless Comm unications,Taoyuan,Taiwan,2001:94～97
10.Ikram M,M organ D.Exploring perm utation inconsistency in blind separation of speech signals in a reverberant environm ent[C].In:Pro-ceedings of IEEE ICASSP,Istanbul,Turkey,2000:1041～1044
11.Ikeda S,M urata N.A m ethod of ICA in tim e-frequency dom ain[C].In:Proceedings of International W orkshop on Independent Com ponent Analysis and Blind Signal Separation(ICA'99),Aussois,France,1999-01:365～371
12.Saruwatari H,Yam ajo H,Takatani T et al.Blind separation and de-convolution of M IM O-FIR system with colored sound inputs using SIM O-M odel-Based ICA[C].In:Proceedings of2003IEEE W orkshop on Statistical Signal Processing(SSP2003),2003-09:421～424
2.Subram aniam S,Petropulu A P,W endt C.Cepstrum-based deconvolu-tion for speech dereverberation[J].IEEE Transactions on Speech and Audio Processing,1996;4(5):392～396
3.Kasprzak W.Blind source deconvolution by hom om orphic filtering in Fourier space[C].In:Proceedings of IAPR ICCVG2002,Zakopane,Poland,2002-09:431～436
4.Lupini A,Solazzi M,Piazza F et al.Frequency domain m ultichannel blind deconvolution using adaptive spline functions[C].In:Proceedings of IEEEINNS-ENNS International Joint Conference on Neural Net-works IJCNN2001,W ashington,USA,2001-07;1:464～468
5.Torkkola K.Blind separation for audio signals-Are we there yet?[C].In:Proceedings of1st International W orkshop on Independent Com-ponent Analysis and Blind Signal Separation,Aussois,France,1999:239～244
6.Sabala I,Cichoki A,Am ari S.Relationships between instantaneous blind source separation and m ultichannel blind deconvolution[C].In:Proceedings of Intl.Joint Conf.on Neural Networks(IJCNN'98),An-chorage,AK,1998-05;1:39～44
7.Yang H H.On-line blind equalization via on-line blind separation[J].Signal Processing,1998;68(3):271～281
8.Zhang Y,Kassam S A.Blind Separation and equalization using frac-tional sam pling of digital com m unication signals[J].Signal Processing,2001;81:2591～2608
9.Enescu M,Zhang Y L,Kassam S A et al.Recursive estim ator for blind M IM O equalization via BSS and fractional sam pling[C].In:Proceedings of Third IEEE Signal Processing W orkshop on Signal Processing Ad-vances in W ireless Comm unications,Taoyuan,Taiwan,2001:94～97
10.Ikram M,M organ D.Exploring perm utation inconsistency in blind separation of speech signals in a reverberant environm ent[C].In:Pro-ceedings of IEEE ICASSP,Istanbul,Turkey,2000:1041～1044
11.Ikeda S,M urata N.A m ethod of ICA in tim e-frequency dom ain[C].In:Proceedings of International W orkshop on Independent Com ponent Analysis and Blind Signal Separation(ICA'99),Aussois,France,1999-01:365～371
12.Saruwatari H,Yam ajo H,Takatani T et al.Blind separation and de-convolution of M IM O-FIR system with colored sound inputs using SIM O-M odel-Based ICA[C].In:Proceedings of2003IEEE W orkshop on Statistical Signal Processing(SSP2003),2003-09:421～424
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心